You should
?¬?nd out what is behind this special naming. Sometimes, the naming can be
done just to satisfy the naming policy for the generated target language,
keeping the generator simpler. If the modelers are not programmers, it is
usually, better to let the generator take care of naming, for example,
removing spaces from entries in models that are used for variable naming.
. Are all modeling concepts needed? If the language includes concepts that are
not applied, those could be removed as they add an extra burden for language
users. Sometimes, during early phases of language creation there is a tendency
to try to covermore of the domain than is actually needed. Finally getting a
chance to master the modeling and code generation, you may go a bit too far.
(2) Model consistency deals with how well the models are kept ???correct.???
. Do the speci?¬?cations follow the language rules? Did we prevent making
false speci?¬?cations or speci?¬?cations that would lead to poor performance?
While metamodel-based rules should be naturally followed in a modern
DSM tool, they are not necessarily speci?¬?ed correctly in the ?¬?rst place. Also
rules that are supported by separate checks might not be followed at all. You
may make checking mandatory by relating it to major modeling actions,
forcing it to be run before versioning the model, or checking models while
generating code, documentation, or other artifacts.
Pages:
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483